Worcestershire Record No. 19 November 2005 p. 63-67

Terry Knight

Survey work in South-east Worcestershire for the Worcestershire Flora Project between 1987 and 2002 produced just under 88,000 records. There were 1179 trips in which over 10 taxa were recorded, 833 of these recorded 50 or more taxa and 200 recorded 100 or more taxa. Computerisation of the survey data has enabled detailed numerical analyses to be carried out, including the calculation of the frequencies of over 600 taxa.

Present Worcestershire and former parts of the county were included in the survey within Ordnance Survey grid squares SP03, SP13, SP14 and part of SP04 as shown in Figure 1. The area comprised 85 whole monads (1km x 1km grid squares) plus 67 part monads with a combined area of 118.6 square kilometres. It is described in more detail in Knight 2005. Trips were spaced out reasonably evenly throughout each year (but principally between 19th March and 18th November) and recorded the date, the monad, the taxa seen and the order in which they were seen.

Groups that were generally under-recorded include submerged water plants; vegetative grasses; subspecies; hybrids; garden escapes; individual species of brambles, dandelions and hawkweeds and the smaller species of willowherbs. During the first few years of the survey some mis-identifications were made resulting in over/under-recording as follows, the Common Name of the plants being generally as given in Stace 1991.

 From the data it is evident that usually only between late May and mid September were more than 100 taxa recorded per trip. As a consequence the trips which have 100 or more taxa would have missed most of the plants that are normally only recorded in the Spring. These include Lesser celandine, Ivy-leaved speedwell, Bulbous buttercup, Thale cress, Sticky mouse-ear, Hairy violet, Wild strawberry and Cuckooflower.

It was decided to base the analysis on data from the first 50 records of the 833 trips. These data contained 617 taxa. For each taxon the percentage of trips in which it was recorded was calculated and taken as the frequency. For comparative purposes, a similar supplementary calculation was carried out for 627 taxa using data from the first 100 records of the 200 trips.

From the total records about 1010 taxa were found with individual species of brambles, dandelions and hawkweeds contributing a further 70 or so taxa. About 210 taxa were single records. The calculated frequency of the first 360 taxa, ranked in order of frequency, from the 833 sets of 50 taxa is given in Table 1. The frequency rank of each of the first 300 or so taxa found from the main calculation and from the supplementary calculation is compared in Figure 2. Two histograms were derived from the two calculations to give the number of taxa in a series of frequency ranges. These are shown in Table 2.

It seems that the calculated frequency depends very much on the number of taxa recorded in each trip. Using sets of 100 records the frequency is significantly higher than when sets of 50 records are used. In contrast, when ranked in order of frequency, the rank of each taxon is much less affected by this as can be seen from Figure 2. (The scatter of points above the general band is mainly from Spring plants under-recorded by the sets of 100 taxa.) However, for analysis on a total of 2000 records (eg 40 sets of 50 taxa), or fewer, the ranking is likely to be fairly unreliable. The histograms in Table 2 suggest there may be a reasonably simple relationship between the number of taxa and the frequency range.

Links to tables & graphs: